diff options
| author | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-06-20 19:23:10 +0300 |
|---|---|---|
| committer | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-06-22 12:02:52 +0300 |
| commit | 729ba46f774a4ba9af48ce6708da653ee80d2296 (patch) | |
| tree | 1a82e24c139d77a8879d85bc8bbb9e3ec8f18ccd /tests/test-backend-ops.cpp | |
| parent | f0325c5826c55bb9796485d49bc971a17735e96a (diff) | |
bitnet(scale in a separate tensor): CPU tweaks
I had ruined TG performance on AVX2 with the last commit.
Was just testing at 8 threads and there we are totally memory
bound. But at 4 threads we had regressed to 41 t/s on the Ryzen7950.
Back to 51 t/s with this commit.
Diffstat (limited to 'tests/test-backend-ops.cpp')
0 files changed, 0 insertions, 0 deletions